Almost five years ago, Abraham Thomas and Tammer Kamel, engineers with decades of experience in the hedge fund industry, set out to create an affordable alternative the $24,000+ per user Bloomberg terminal. “As an analyst, you spend 70-80% of your time collecting and cleaning data,” explained Abraham Thomas, one of Quandl’s founders, in a recent interview. “Bloomberg is great for clean, curated data,” he continued, “but if you’re not one of those lucky few, you’re out of luck. And the irony is that there’s no shortage of data out there, it’s just usually in a PDF when you find it!”
Quandl hopes to create value by disintermediating industry incumbents and connecting data owners directly with data users.
The data services industry is highly consolidated, and companies have long organized as resellers that either a) hire analysts to build datasets from open sources, b) purchase and resell structured datasets from smaller vendors or c) arrange cross-licensing agreements with other industry participants. Before much of this data became available through websites like the SEC’s EDGAR database, Reuters bought one share in every company and typed in the numbers from shareholder reports. Much of this expensive data is just collected from government websites and cleaned up. In this sense, traditional data services companies act as intermediaries between those who own data and those who use data.
Incumbents have benefited from scale for a number of good reasons. First, by bringing together disparate data structures into a uniform schema, Bloomberg and Reuters take advantage of aggregation effects, in which selling products together make the products more valuable. Second, in order to effectively price discriminate in a zero-marginal costs digital industry, traditional firms have relied heavily on bundling data products into a single annual price tag. Third, cross-licensing agreements have led an exclusive club in which you’d better have a lot of data to offer if you hope to negotiate a cross licensing agreement with the big players.
Quandl solved the classic platform liquidity problem by aggregating enough open data to attract users.
Quandl’s founders recognized that they faced the chicken-and-egg problem that all two-sided market businesses face: they had to get enough users to be valuable as a distribution platform for data vendors, while they had to get enough data on the platform to get enough users in the first place.
Abraham and Tammer decided to attack the user side of the problem first by building enough liquidity on the data side of the market by themselves. They spent more than two years scraping open datasets and posting them on Quandl. Where possible, they would build scripts to automate the process. They also decided to make the data available for download or through an API. Last but not least, Quandl used a freemium model to acquire users, with all the open data they had cleaned and posted accessible for no charge.
By late 2014, Quandl had reached enough users to finally solve the liquidity problem. With over 10,000 users and millions of downloads per month, the company began to receive phone calls from smaller data vendors eager to sell their data through the Quandl platform. When I asked for an example, of the type of data vendor, Abraham mentioned Zach’s Investment Research, a Chicago-based research firm that has been known since the 1970s for its analyst surveys and cash-flow-adjusted financials.
Like many platform businesses, Quandl will capture value by charging data vendors a commission for the data sold through their platform.
As of March 2015, Quandl had a dozen data vendors selling “premium” datasets through their platform, and a dozen more in the pipeline. “We’re like an App store for data,” Abraham quipped. I don’t know what their commission rates are, but a lot of the premium datasets on quandl.com are going for north of $5k per license.
Interview with Abraham Thomas on 13 March 2015